Multiple Imputation for Combined-Survey Estimation with Incomplete Regressors in One But Not Both Surveys

39 Pages Posted: 4 Jan 2012

See all articles by Michael S. Rendall

Michael S. Rendall

RAND Corporation

Bonnie Ghosh-Dastidar

affiliation not provided to SSRN

Margaret Weden

RAND Corporation

Zafar Nazarov

RAND Corporation

Date Written: October 24, 2011

Abstract

Within-survey multiple imputation (MI) methods are adapted to pooled-survey regression estimation where one survey has a larger set of regressors but fewer observations than the other. This adaption is achieved through: (1) larger numbers of imputations to compensate for the higher fraction of missing values; (2) model-fit statistics to check the assumption that the two surveys sample from a common universe; and (3) specificying the analysis model completely from variables present in the survey with the larger set of regressors, thereby excluding variables never jointly observed. In contrast to the typical within-survey MI context, cross-survey missingness is monotonic and easily satisfies the Missing At Random (MAR) assumption needed for unbiased MI. Large efficiency gains in estimates of coefficients for variables in common between the surveys are demonstrated in an application to sociodemographic differences in the risk of experiencing a disabling occupational injury estimated from two nationally-representative panel surveys.

Suggested Citation

Rendall, Michael S. and Ghosh-Dastidar, Bonnie and Weden, Margaret and Nazarov, Zafar, Multiple Imputation for Combined-Survey Estimation with Incomplete Regressors in One But Not Both Surveys (October 24, 2011). RAND Working Paper Series No. WR-887-1, Available at SSRN: https://ssrn.com/abstract=1979715 or http://dx.doi.org/10.2139/ssrn.1979715

Michael S. Rendall (Contact Author)

RAND Corporation ( email )

1776 Main Street
P.O. Box 2138
Santa Monica, CA 90407-2138
United States

Bonnie Ghosh-Dastidar

affiliation not provided to SSRN ( email )

Margaret Weden

RAND Corporation ( email )

1776 Main Street
P.O. Box 2138
Santa Monica, CA 90407-2138
United States

Zafar Nazarov

RAND Corporation ( email )

1776 Main Street
P.O. Box 2138
Santa Monica, CA 90407-2138
United States

Do you have negative results from your research you’d like to share?

Paper statistics

Downloads
57
Abstract Views
513
Rank
659,215
PlumX Metrics